Why COVID-19 and other infectious diseases do not automatically become milder

By Martin Michaelis and Mark Wass. Many argue that the COVID-19 restrictions should be eased more quickly because infectious diseases become milder over time. However, this is actually a misconception.

An often communicated belief is that infectious diseases automatically become milder over time. This idea stems from computational modelling studies, which show that the best evolutionary strategy for an infectious agent is not to cause any damage to its host. Moreover, increasing immune protection from previous infections (and vaccinations) is anticipated to reduce the risk of severe disease from future infections. Although these are reasonable assumptions, this is not always what happens.

Smallpox, a disease caused by Variola viruses, was found in mummies from the 3rd century B.C. Nevertheless, there is no evidence that smallpox had become milder by 1977, when a global vaccination programme eradicated the disease. In the last 100 years of its existence, smallpox is estimated to have killed half a billion people.

Similarly, the plague, which is caused by the bacterium Yersinia pestis, does not appear to have become less severe over the millennia. Yersinia pestis was found in a Swedish tomb from 3,000 B.C. and may have been associated with a rapid collapse of human populations at the time (something known as the Neolithic decline). Large plague outbreaks have been described since Roman times.

At the beginning of the 20th century, plague epidemics still occurred in North America. Then, Yersinia pestis was identified as the infectious agent causing the plague after transmission from rodents to humans by fleas and other insects. Based on this understanding, improved hygiene and pest control measures, plague vaccines, and antibiotics resulted in a strong decline of plague cases. Today, there are about 600 cases of plague reported globally per year, and without treatment the mortality rates remain as high as ever at about 70%. Even with antibiotic therapy, about 11% of plague patients die in a developed country like the US.

Another example is influenza, commonly called ‘the flu’. Evidence suggests that an influenza outbreak already happened in China in 6,000 B.C. Nevertheless, the Spanish Flu, caused by a so-called H1N1 influenza A virus in 1918-1920, was the deadliest pandemic of modern times causing an estimated 17-100 million deaths.

It has also just been shown that the death risk associated with the B.1.1.7 variant of SARS-CoV-2, the coronavirus that causes COVID-19, is about 60% higher than that associated with previous variants. Moreover, the increasing spread of the South African B.1.351 variant was associated with an increase of the in-hospital mortality by 20%. This shows that some of the novel variants, which are currently emerging, are already deadlier than the original SARS-CoV-2 variants.

Finally, there are even infectious diseases in which repeated infections increase the risk of severe disease. The most prominent example of this is Dengue fever, which is caused in tropical regions by the Dengue virus after transmission from mosquitoes. There are different Dengue virus types and infection with one type can render individuals particularly vulnerable to severe, life-threatening complications from other types, a phenomenon called antibody-dependent enhancement.

Although mathematical modelling studies suggest that it is in the interest of an infectious agent not to cause disease and to reduce its virulence over time, this is not always what happens. Reasons for this include that a reduction in virulence can mean that a pathogen also loses its fitness and contagiousness in a real-life setting. Moreover, pathogens like influenza viruses (and most probably also SARS-CoV-2), are variable enough to develop new variants that can bypass pre-existing immunity caused by previous infections and vaccinations. In the worst case, variants may emerge that cause severe disease particularly in those who have previously been infected with another variant.